He's the Best

If you're a diehard baseball fan, you're familiar with some of the most in-depth statistics ever devised. There are absurd numbers now (line-drive percentage, batting average on balls in play, etc.) that provide an infinitely more accurate assessment of individual players than what we had a decade ago.

In football, though, that statistical fanaticism has never really caught on. I don't think anyone could tell you how pass efficiency is calculated or what it really tells you about a QB, and even numbers that apply in multiple sports (value over replacement player, for example) have been widely ignored.

Fortunately, thanks to the magic of the interwebs, we have sites like Smart Football and Football Outsiders that are championing the statistical revolution, allowing us to compare players and statistics on even terms in a way that was never before possible.

When I looked at some rushing numbers from last year and came to the realization that Cal's Jahvid Best averaged an absurd 8.14 (!!!) yards per carry, I figured that at least part of that total came from a low number of carries, a weak schedule ... I dunno, something. Basically, that number just seemed too good to be true.

But was it? Bill Connolly at Football Outsiders posted an article last week in which he essentially determined a running back's value over replacement player, or what he called "points over expected." The math is a little complicated, but the general idea is to combine yards per carry (which can be skewed by a low number of carries ) with total yards (which can be skewed with a high number of carries) and determine a value for each running back compared with what a hypothetically average player would produce in the same circumstances. A "plus" number is good (meaning the player is better than an average replacement would be), while a "minus" number is bad.

The results:

Yeah, Best was pretty good no matter how you break it down. In fact, according to POE, he was easily the best running back in the country. His POE value of over 40 means that he produced at a 40% higher rate than the average replacement player would have in the same circumstances (the same offense against the same defenses, basically), which is pretty freakin' impressive.

Does that mean he'll be the best again this year? Well ... probably, although there's one guy in particular who I think could give him a run for his money.

Three more of the top five (in POE) running backs from last year also return -- LeGarrette Blount at Oregon, Jonathan Dwyer at Georgia Tech and Kendall Hunter at Oklahoma State -- and the name that's both surprising and intriguing is Blount, who platooned with Jeremiah Johnson but finished with the third-highest "points per play" value (essentially yards per carry) in the country. As the full-time starter in what should be an explosive Ducks offense under Chip Kelly, Blount could end up with some ridiculous numbers this year. The only thing that might keep him from putting up 1,500 yards: Jeremiah Masoli, who finished as the fourth-rated rushing quarterback in the country. In other words, Oregon's offense = terrifying.

Back to Best, though, I'll be a little surprised if he doesn't make a run at 2,000 yards this season. The loss of All-American center Alex Mack will hurt some, and the line has been shuffled a bit in Mack's absence, but both tackles (Mitchell Schwartz and Mike Tepper) should garner All-Pac-10 consideration and both Chris Guarnero at center and Chet Teofilo at guard have starting experience.

As much as he loves the passing game, Jeff Tedford has to know that Cal's best chance to win the Pac-10 this year is to put the ball in Best's hands as often as possible. And even if his per-carry average drops a full yard to a still-ridiculous 7.14, he would need only 280 carries (or 21.4 per game over a 13-game season, including a bowl game) to reach 2,000.

As Keyshawn Johnson would say, just give him the damn ball.